In working on DM-19189, I have been looking at the S/N distributions of the various fluxes  (raw instrumental and calibrated) from single frame processing (where here S/N = flux/fluxErr from the catalog).  The S/N distributions look as expected for the raw instrumental fluxes, the calibrated values based on application of the photoCalib of jointcal outputs, and the meas_mosaic solution using the "old style" calibration (i.e. using meas_mosaic's own applyMosaicResultsExposure() function – which uses the persisted fcr dataset to apply the calibration).  However, the S/N distribution when applying the photoCalib from meas_mosaic is not at all as expected (plots to be attached). Look further into this pathology and what might be causing it.